
Eurographics/ACM SIGGRAPH Symposium on Computer Animation (2004) R. Boulic, D. K. Pai (Editors) Breathe Easy: Model and control of simulated respiration for animation Victor B. Zordan, Bhrigu Celly, Bill Chiu, and Paul C. DiLorenzoy University of California Riverside Abstract Animation of the breath has been largely ignored by the graphics community, even though it is a signature move- ment of the human body and an indicator for lifelike motion. In this paper, we present an anatomically inspired, physically based model of the human torso for the visual simulation of respiration using a mixed system of rigid and deformable parts. This novel composition of anatomical components is necessary to capture the key charac- teristics of breathing motion visible in the human trunk because the movement is generated fundamentally through the combination of both rigid bone and soft tissue. We propose a simple anatomically meaningful muscle ele- ment based on springs, which is used throughout both actively to drive the motion of the ribs and diaphragm and passively for other muscles like those of the abdomen. In addition, we introduce a straightforward method for preserving incompressible volume in deformable bodies to use in approximating the motion of the gut related to breath. Through the careful construction of this anatomically based torso, control for respiration becomes the generation of periodic contraction signals for a minimal set of two muscle groups. We show the flexibility of our approach through the animation of several breathing styles using our system and we verify our results through video and analytical comparisons. Categories and Subject Descriptors (according to ACM CCS): I.3.5 [Computer Graphics]: Three Dimensional Graphics and Realism:Animation; I.3.7 [Computer Graphics]: Computational Geometry and Object Modeling, Physically-Based Modeling 1. Introduction Breathing is a critical body function and its motion is a tell- To capture the many and complex interactions that are tale signature for the living. At rest or asleep, the involuntary seen between the variety of components related to breath, movement of the trunk is predominantly driven by the com- a physical anatomical model is an obvious choice. This ap- plex biological process of the breath. However, in animation, proach is superior to describing the motion procedurally be- respiration and its deformation of the torso have remained cause the functions of breath interplay in complex ways and stylistic and are often overly simplified or ignored entirely. are difficult to explain heuristically because of the mixing To create a believable moving body, especially within and of deformation and rigid body motion. The movement as- around the torso, and to visually bring a character to life, the sociated with breath could be isolated during capture with movement caused by breathing is invaluable. In this paper, data-driven skin deformation approaches [ACP02, SMP03] we present a system which mimics the biological functions but a physical simulation will allow finer control over the of respiration through a simple, physically based, anatomi- subtleties of the movement and can encapsulate a range of cally inspired simulation with the goal of synthesizing the breathing behaviors in a single representation that can gen- motion associated with human breath for computer anima- erate novel motion immediately without the need for addi- tion. tional recording. To create the overall visual effects found in the motion y e-mail: vbz j bcelly j bill j [email protected] of breath, we propose a simple composite simulation which c The Eurographics Association 2004. Zordan, Celly, Chiu, DiLorenzo / Breathe Easy: Model and control of simulated respiration for animation Dinosaur [Dis00]. Physically based approaches for skeletal muscles include the work of Chen and Zeltzer [CZ92] who use a biologically based muscle model to generate proper muscle force and Teran and colleagues who use a finite vol- ume method (FVM) to create a continuous internal-tension based muscle simulation [TBHF03]. Both show results of deformation on the muscles systems of a single limb. In addi- tion, Nedel and Thalmann propose the use of a spring-mass system as an alternative for real-time applications [NT98]. Closest to our own efforts are respiration models of Kaye et al. [KMJ97] who animate deformable lungs for clinical applications based on a model built from CT scans and sim- plified cardiopulmonary mechanics and the constraint-based solver of Promayon et al. [PBP97] which models the defor- mation of the abdomen during calm breath. However to the best of our knowledge, ours is the first work to investigate the animation of breath by simulating the motion of both the ribcage and gut. Figure 1: Full skeletal model in lotus position, image from Our system combines a custom deformable simula- an animation of calm breathing. This figure shows the com- tion system, that preserves volume based on pressure, bination of rigid-body bones and deformable gut as well with an available rigid-body dynamics solver, Open Dy- as secondary elements (bones) added for the shoulders and namics Engine [Smi03]. Since the pioneer work by Ter- arms. zopoulos et al. [TPBF87, TF88] introducing the use of differential equations to animate deformation, numer- ous researchers have suggested techniques for interac- tive and multi-resolution deformable simulation, includ- combines rigid-body dynamics with elastically deformable ing [JP99, DDCB01, CGC∗02, GKS02, MDM∗02]. In gen- bodies. The simulation uses spring-based muscles to esti- eral, exact volume preservation is not guaranteed by a given mate forces that pull and deform the bodies and estimated deformation system, though it may afford a structurally sup- pressure forces to preserve the volume of the deformable ported volume, for example, by constructing objects us- components. Because the human body incorporates soft, ing 3D tetrahedrons elements, as Muller and his colleagues deformable organs and muscles with (mostly) rigid bones, demonstrate [MDM∗02]. Deformation with explicit vol- approaches which capture only one form of motion, de- ume preservation has been managed in fewer cases: several formable or rigid-body, are insufficient and will lead to ei- suggest techniques using constraint solvers and optimiza- ther computational limitations or a lack of flexibility. The tion [PB88, RSB96, PBP97]; Cani-Gascuel and Desbrun use use of rigid bodies and spring-based systems has appeared in implicit surfaces and add a translation function to the surface numerous research and commercial arenas associated with displacement to account for changes in volume [CGD97]; graphics, but few have discussed the interaction of such and Teran and colleagues allow for preservation through a systems [OZH00, BW97]. Further, none to our knowledge volumetric term added to the internal tension of a muscle have proposed a system of like scale which seamlessly com- modeled with FVM [TBHF03]. Also, a real-time approach bines such components. While we choose individual simula- is offered by Stahl and colleagues for simulation of tissue tions of the base components that are each simple and well- volume with a constrained “bag of particles” [SET02]. understood, our approach is novel in its use and integration of these components and affords our top-level goal to faith- While rigid-body dynamics is well-understood and de- fully recreate the complex motion associated with breath. scribed in many texts, control for motion has been the fo- cus of most rigid-body related papers found in the liter- ature for computer graphics. Though no truly general so- 2. Background lutions for control have been offered to date, most have Visual and physical simulation of synthetic anatomical simplified muscle activation to torque-generated actuation. muscles has been described for several applications re- Because we use direct muscle force activation in lieu of lated to modeling and animation, for example in the torque-driven motion, we save remarking on these many ef- head and face [PB81, Wat87, TK90, LTW95, KHYS02], forts for brevity. Techniques using force-based controllers the hand [GMTT89, AHS03], and for skeletal mus- for simulated behaviors are much less common, two exam- cles [CZ92, SPCM97, WG97, NT98, TBHF03]. Visual sim- ples being the spring-actuated controllers employed to an- ulation of skeletal muscles has been approached procedu- imate flexible models for snakes and fishes [Mil88, TT94]. rally through heuristic shape changes made in response to To create behaviors for slithering and swimming, control bone movement [SPCM97, WG97]. These examples model systems are constructed with hand-tuned input parame- the change in shape of a muscle through geometric muscle ters for sinusoids that move the body through coordinated bellies that stretch and deform based on length. Such pro- forces. Follow-up work shows that optimization is useful in cedural techniques have been adopted in the entertainment generating these control parameters automatically [GT95]. industry and used extensively for movies such as Disney's Other related approaches introduce alternative methods for c The Eurographics Association 2004. Zordan, Celly, Chiu, DiLorenzo / Breathe Easy: Model and control of simulated respiration for animation controlling free-form deformations and mass-spring lat- tices [WW90, FvdPT97, CMN97]. Data-driven methods offer alternative approaches to physically based models for creating realistic motion. Most recently, capture technologies like
Details
-
File Typepdf
-
Upload Time-
-
Content LanguagesEnglish
-
Upload UserAnonymous/Not logged-in
-
File Pages9 Page
-
File Size-